RazorBrain's Go Journal


How to Work Life & Death Problems by RazorBrain
March 11, 2009, 05:00
Filed under: Go Study, Tsumego, Video/Screencast | Tags: ,

One way to improve your game dramatically is to regularly work life and death problems. But many beginners are put off by this because they lack a workable approach to tackling challenging problems. I have found that working life and death problems requires patience more than anything else. Yes, a trained eye is useful in a game, but we develop the ‘intuition,’ if you will, that a trained eye depends upon by working lot’s of life and death problems both in and out of games. This post then is all about process and patience for new players.

We develop the ‘intuition’ that a trained eye depends upon by working lot’s of life and death problems.

When you don’t have the experience to ‘see the obvious move’ what do you do? The answer is that you let process and patience do their jobs until the answer is clear. To begin, let’s take a tried and true method that I first read about in a book titled “Tesuji” by James Davies. We’ll apply patience, perhaps throw in a bit of stubbornness and we’ll learn how to effectively work life & death problems. The method or process is easy but can seem hard when we try to apply it to problems that stretch us. But it works! stick with it.

Here’s the method: select the move you think is the right one and play it in your head, then respond for your opponent with the best response you can think of. Then continue making moves for both you and your opponent until you prove that your moves are correct or incorrect. They are correct if you are able to kill or live depending on the objective of the problem despite any attempt by your opponent to foil your plans.

Something is incorrect if one of your opponent’s responses causes you to miss your objective. If you see that one of your moves is incorrect, back up to that move and try another one, then again respond for your opponent just as you did before. If you see that your moves seem to work, then back up again and this time play a different response for your opponent and play out the problem again. Continue to do this for both you and your opponent until you have exhausted all of the possible options.

Discipline and concentration will win everytime over a good guess. So, while you might not be able to work out problems that are way above your level, you should be able to patiently work out problems rated at your strength or slightly higher without your brain tapping out and begging for a guess. The video below shows an example of this method at work. The example problem is rated at 20 kyu. So, if you’re between 30 and 15 kyu, you should find this video helpful, especially if you are new to working life and death problems.

If you’d like to see other video tutorials on this topic or other topics, leave me a comment and I’ll add your request to my list of blog projects.

Advertisement

8 Comments so far
Leave a comment

Hello! Interesting go blog!! :)

I’m a spanish boy, 6k kgs now, I’m improving. No time, but i’m trying to do my best.

Visit me at…

http://elcaminoaldan.wordpress.com

In spanish, some post in English. And a lot of material to share. :)

Comment by balddy

Thanks for the problem! I am 11k on KGS and this was eye-opening for me!

Comment by Terrence (metaperl)

[...] Posts Tsumego – Intermediate #4Tsumego Walkthrough – 5kHow to Work Life & Death ProblemsThe Seven Deadly Sins (of Go)Tsumego 25 Kyu – Women and Children [...]

Pingback by Tsumego 25 Kyu - Lose the Tunnel Vision « RazorBrain’s Go Journal

[...] Top Posts Game Review for CreedFox (16 kyu)The Seven Deadly Sins (of Go)Tsumego 25 Kyu – Gotta Kill Me Some StonesTetsuki for iPhoneHow to Work Life & Death Problems [...]

Pingback by Tsumego 25 Kyu - Gotta Kill Me Some Stones « RazorBrain’s Go Journal

Thanks for the video!
It took me relatively long to find A16, just as you said, the solution is rather not obvious. (at least for a player like me)
Solving tsumego, I found out that it is quite hard to keep count on the liberties of every group. In this case, the shortage of liberties for white kills the group, which is why the move at A16 and the following sequence is not the first we think about.
Keep up the good work :)

Comment by Robo PK Team 2009

I think I was misunderstood a little.

The systematic process is typically done with tsumego — the top/down left/right approach was in the tsumego context. What I’m curious about is tips beyond a systematic search for a /human/ to do when trying to solve tsumego (systematic search being what you described in the video…ie. try all options until you’ve exhaustively shown what the best move is).

This, of course, only can really be used locally because of the time it takes, and I realize that intuition should be used more-so than trying all of the plays during the real game. I’m not looking for better heuristics for a computer Go AI, but better heuristics for a human to solve tsumego. For example, do you try to rule out as many of the initial moves you can perform first? (ie. by seeing the possible play-outs from what looks like a bad move) …or do you try the best moves first? (ie. try to hit the vital points and play it out)

A systematic search takes patience, and practice is the best way to learn that, but it’s the extra tidbits of information that is precious to learn.

Thanks. Cheers

Comment by haz

Christian, thanks for the comment. I’m not a computer go programer, but I suspect the methods you enquire about are used in go programming. However, for the humans out ther, I’ve never heard of this type of method being employed. I would that this is so for the following reasons. First, as you gain strength, you experience will enhance your subconsious processing of these problems. Many call this intuition and perhaps part of it is. But to suggest that we approach every problem from a strictly left-to right, or top-to-bottom approach and try every option is not realistic for live games, though it works pretty well for tsumego.

Don’t get me wrong though. You can use a method like this, but you won’t have to try every option as you get stronger. However, as you play more you’ll find that vital points of certain shape scream at you for attention. So, you will find it difficult to stick to a right-to-left method once this happens for you. That is why we say to pick what you think the best move is and start from there. Ha, with tsumego, the authors try to make the problems a bit tricky, but in real games sometimes the most obvious move works. So, why waste the time trying all of the others. The practice of tsumego just helps our brains adjust to the process and comes in very handy when the going gets tough and you have to methodcally work though it. So, for the tough times, a right-to-left or top-to-bottom approach might help people keep track of what they’ve already tried.

I’m glad you enjoyed the post. I plan on starting to post 25 kyu problems once a week beginning next week. I’ll be doing other levels too well as game reviews. Let me know whatyou’d like to see and I’ll try to do it.

Comment by RazorBrain

It may just be my comp sci background, but the process you describe screams obvious to me — its just the simple backtracking algorithm that I would suspect any player would apply to systematically solve a problem.

What I’m curious about (as a beginner go player), is heuristics that I can apply mentally while going through the play tree in my mind. One issue is remembering everything you’ve tried before, and a precedence on how you check things can simplify this greatly — for example if I always test plays from top->bottom, left->right, then I don’t need to remember /all/ the things I’ve tried…just the move leading up to the position I’m trying to visualize.

If there are any tricks to knowing when to give up, then less systematic search would need to be done. Also, if there are any tricks to equate different options (ie. plays A-B-C and C-B-A end up in the same spot), then that would save time as well.

Any chance of these extra tidbits being included in future posts?

Thanks. Keep up the great work.

Comment by Christian Muise




Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Connecting to %s



Follow

Get every new post delivered to your Inbox.